FARE-20B is a multi-task evaluator fine-tuned from gpt-oss-20B. It’s trained on a broad, multi-domain dataset using rejection-sampled SFT and can handle tasks like pairwise comparison, step-level scoring, reference-based and reference-free verification, and single-score assessment.
Fine-tuningDocs | FARE-20B can be customized with your data to improve responses. Fireworks uses LoRA to efficiently train and deploy your personalized model |
On-demand DeploymentDocs | On-demand deployments allow you to use FARE-20B on dedicated GPUs with Fireworks' high-performance serving stack with high reliability and no rate limits. |
